Coding
Like-for-like- GPT-5.4
- 57.7
- Hy4 preview
- 65.7
- Weighted basis
- 1 vs 1 rows
- Reading
- Hy4 preview leads
GPT-5.4 vs Hy4 preview
Updated August 28, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Hy4 preview has the higher public score estimate, 79.16 versus 73.56, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
9 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Coding work
Code generation, repair, and software-engineering tasks
Hy4 preview
Hy4 preview leads on the same 1 weighted benchmark row.
Confidence: limited
Long documents
Prompts that approach the documented context limit
GPT-5.4
GPT-5.4 has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
28 GPT-5.4 only
9 Shared
19 Hy4 preview only
- Shared results
- 9
- GPT-5.4 only
- 28
- Hy4 preview only
- 19
- Like-for-like categories
- 2 / 8
1 category uses different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.
Knowledge
Like-for-like- GPT-5.4
- 57.6
- Hy4 preview
- 60.4
- Weighted basis
- 2 vs 2 rows
- Reading
- Hy4 preview leads
Multimodal
Directional only- GPT-5.4
- 73.2
- Hy4 preview
- 66.2
- Weighted basis
- 3 vs 1 rows
- Reading
- Directional only
Agentic
Not comparable- GPT-5.4
- 77.2
- Hy4 preview
- Not measured
- Weighted basis
- 3 vs 0 rows
- Reading
- Not comparable
Reasoning
Not comparable- GPT-5.4
- 74.0
- Hy4 preview
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
Math
Not comparable- GPT-5.4
- 42.5
- Hy4 preview
- Not measured
- Weighted basis
- 2 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- GPT-5.4
- Not measured
- Hy4 preview
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Instruction following
Not comparable- GPT-5.4
- Not measured
- Hy4 preview
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | GPT-5.4 | Hy4 preview | Weighted basis | Reading |
|---|---|---|---|---|
| Coding | 57.7 | 65.7 | Like-for-like1 vs 1 rows | Hy4 preview leads |
| Knowledge | 57.6 | 60.4 | Like-for-like2 vs 2 rows | Hy4 preview leads |
| Multimodal | 73.2 | 66.2 | Directional only3 vs 1 rows | Directional only |
| Agentic | 77.2 | Not measured | Not comparable3 vs 0 rows | Not comparable |
| Reasoning | 74.0 | Not measured | Not comparable1 vs 0 rows | Not comparable |
| Math | 42.5 | Not measured | Not comparable2 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Instruction following | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
OfficeQA Pro
Multimodal
SWE-bench Pro
Coding
HLE
Knowledge
GPQA
Knowledge
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- GPT-5.4
- $0.01
- Fits in one request
- Hy4 preview
- Self-hosted; infrastructure cost varies
- Fits in one request
Hy4 preview has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- GPT-5.4
- $0.17
- Fits in one request
- Hy4 preview
- Self-hosted; infrastructure cost varies
- Fits in one request
Hy4 preview has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- GPT-5.4
- $0.25
- Fits in one request
- Hy4 preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
Hy4 preview has no comparable published API token rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationGPT-5.4Hy4 preview
Context window
Maximum documented context; output-token limits may be lower.
GPT-5.4
1.05M
OpenAI pricingHy4 preview
API model ID
GPT-5.4
gpt-5.4
OpenAI pricingHy4 preview
tencent/Hy4-preview
Tencent Hy4 preview model cardCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
GPT-5.4
$0.25 per 1M cached input tokens
OpenAI pricingHy4 preview
No comparable hosted API rate
Tencent Hy4 preview model cardDocumented inputs
GPT-5.4
Not sourced
Hy4 preview
Not sourced
Documented outputs
GPT-5.4
Not sourced
Hy4 preview
Not sourced
Provider availability
GPT-5.4
Not sourced
Hy4 preview
Not sourced
Reasoning profile
GPT-5.4
Reasoning
Hy4 preview
Reasoning
Weight access
GPT-5.4
Proprietary
Hy4 preview
Open Weight
License
GPT-5.4
Proprietary
Hy4 preview
Open Weight
Release date
GPT-5.4
2026-03-05
Hy4 preview
2026-08-28
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- Hy4 preview has the higher public score estimate, 79.16 versus 73.56, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- GPT-5.4 has the larger documented window (1.05M).
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence56 rows
Agentic
Terminal-Bench 2.0
GPT-5.475.1%SourceHy4 preview—Not directly comparable
CyberGym
GPT-5.479.0%SourceHy4 preview78.4%SourceGPT-5.4 leads this result
BrowseComp
GPT-5.482.7%SourceHy4 preview—Not directly comparable
OSWorld-Verified
GPT-5.475%SourceHy4 preview—Not directly comparable
MCP Atlas
GPT-5.470.6%SourceHy4 preview83.7%SourceHy4 preview leads this result
Toolathlon
GPT-5.454.6%SourceHy4 preview—Not directly comparable
τ²-bench results
GPT-5.498.9%SourceHy4 preview—Not directly comparable
Claw-Eval
GPT-5.460.3%SourceHy4 preview—Not directly comparable
DeepSearchQA
GPT-5.473.6%SourceHy4 preview—Not directly comparable
Gert Labs
GPT-5.464.89%SourceHy4 preview—Not directly comparable
ResearchClawBench
GPT-5.415.3%SourceHy4 preview—Not directly comparable
JobBench
GPT-5.438.9%SourceHy4 preview61.7%SourceHy4 preview leads this result
ExploitGym
GPT-5.46.0%SourceHy4 preview—Not directly comparable
Terminal-Bench 2.1
GPT-5.4—Hy4 preview85.4%SourceNot directly comparable
WideResearch
GPT-5.4—Hy4 preview83.9%SourceNot directly comparable
DRACO
GPT-5.4—Hy4 preview77.2%SourceNot directly comparable
Toolathlon-Verified
GPT-5.4—Hy4 preview74.1%SourceNot directly comparable
APEX-Agents
GPT-5.4—Hy4 preview37.1%SourceNot directly comparable
skillsBench
GPT-5.4—Hy4 preview62.9%SourceNot directly comparable
Agents' Last Exam
GPT-5.4—Hy4 preview22.8%SourceNot directly comparable
AutomationBench
GPT-5.4—Hy4 preview32.1%SourceNot directly comparable
BankerToolBench
GPT-5.4—Hy4 preview78.6%SourceNot directly comparable
HLE w/ tools
GPT-5.4—Hy4 preview55.4%SourceNot directly comparable
Coding
LiveCodeBench Pro
GPT-5.487.5%SourceHy4 preview—Not directly comparable
SWE-bench Pro
GPT-5.457.7%SourceHy4 preview65.7%SourceHy4 preview leads this result
React Native Evals
GPT-5.485.3%SourceHy4 preview—Not directly comparable
Vibe Code Bench
GPT-5.467.42%SourceHy4 preview—Not directly comparable
Terminal-Bench 2.1
GPT-5.4—Hy4 preview85.4%SourceNot directly comparable
SWE Multilingual
GPT-5.4—Hy4 preview82.9%SourceNot directly comparable
deepSwe
GPT-5.4—Hy4 preview64.3%SourceNot directly comparable
NL2Repo
GPT-5.4—Hy4 preview58.9%SourceNot directly comparable
ProgramBench
GPT-5.4—Hy4 preview17.5%SourceNot directly comparable
PostTrain Bench
GPT-5.4—Hy4 preview35.6%SourceNot directly comparable
sweMarathon
GPT-5.4—Hy4 preview31.9%SourceNot directly comparable
Reasoning
Knowledge
GPQA
GPT-5.492.8%SourceHy4 preview92.3%SourceGPT-5.4 leads this result
HLE
GPT-5.452.1%SourceHy4 preview55.4%SourceHy4 preview leads this result
HLE w/o tools
GPT-5.439.8%SourceHy4 preview43.4%SourceHy4 preview leads this result
GPQA-D
GPT-5.492.8%SourceHy4 preview92.3%SourceGPT-5.4 leads this result
HealthBench Hard
GPT-5.440.1%SourceHy4 preview—Not directly comparable
MedXpertQA (Text)
GPT-5.459.6%SourceHy4 preview—Not directly comparable
HealthBench Professional
GPT-5.448.1%SourceHy4 preview—Not directly comparable
Math
Multimodal
MMMU-Pro
GPT-5.481.2%SourceHy4 preview—Not directly comparable
OfficeQA Pro
GPT-5.453.2%SourceHy4 preview66.2%SourceHy4 preview leads this result
MMMU-Pro w/ Python
GPT-5.482.1%SourceHy4 preview—Not directly comparable
CharXiv
GPT-5.482.8%SourceHy4 preview—Not directly comparable
ERQA
GPT-5.465.4%SourceHy4 preview—Not directly comparable
SimpleVQA
GPT-5.461.1%SourceHy4 preview—Not directly comparable
ScreenSpot Pro
GPT-5.485.4%SourceHy4 preview—Not directly comparable
ZeroBench
GPT-5.441.0%SourceHy4 preview—Not directly comparable
MedXpertQA (MM)
GPT-5.477.1%SourceHy4 preview—Not directly comparable
Frequently asked questions
Which is better, GPT-5.4 or Hy4 preview?
Hy4 preview has the higher public score estimate, 79.16 versus 73.56, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, GPT-5.4 or Hy4 preview?
Hy4 preview leads the like-for-like coding comparison across 1 shared weighted benchmark row.
Which is better for agentic tasks, GPT-5.4 or Hy4 preview?
The published evidence does not provide a shared weighted agentic tasks basis for both models, so BenchLM does not name a agentic tasks winner.
Which costs less, GPT-5.4 or Hy4 preview?
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Which has the larger context window, GPT-5.4 or Hy4 preview?
GPT-5.4 has the larger documented context window: 1.05M, compared with 1M.
Related comparisons
Last updated August 28, 2026
Watch GPT-5.4 vs Hy4 preview
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.